sign in sign up
<<
Home , Citrullination, LSm

Diabetes Volume 64, February 2015 573

Dieter Rondas,1 Inne Crèvecoeur,1 Wannes D’Hertog,1 Gabriela Bomfim Ferreira,1 An Staes,2,3 Abhishek D. Garg,4 Decio L. Eizirik,5 Patrizia Agostinis,4 Kris Gevaert,2,3 Lut Overbergh,1 and Chantal Mathieu1

Citrullinated Glucose- Regulated 78 Is an Autoantigen in Type 1 Diabetes

Diabetes 2015;64:573–586 | DOI: 10.2337/db14-0621

Posttranslational modifications of self- play a sub- the autoimmune assault progresses, with – fi stantial role in the initiation or propagation of the autoim- appearing against several non b-cell-speci c autoantigens, TRANSPLANTATION AND IMMUNOLOGY mune attack in several autoimmune diseases, but their such as GAD65 (3), islet 2 (IA2) (4), heat shock contribution to type 1 diabetesisonlyrecentlyemerging.In protein 60 (HSP60) (5), and chromogranin A (ChgA) (6). the current study, we demonstrate that inflammatory stress, During insulitis, local production of inflammatory media- induced by the cytokines interleukin-1b and interferon-g, tors, such as the cytokines interleukin (IL)-1b and interferon-g leads to citrullination of GRP78 in b-cells. This is coupled (IFNg), triggers b-cell oxidative and endoplasmic reticulum with translocation of this endoplasmic reticulum (ER) stress. These, and other signals, may lead to alternative to the b-cell plasma membrane and subsequent secretion. splicing and misfolding of b-cellproteinsaswellasposttrans- Importantly, expression and activity of peptidylarginine lational modifications (PTMs) (7–9).Inotherautoimmune deiminase 2, one of the five enzymes responsible for citrul- diseases, like (RA), , lination and a candidate gene for type 1 diabetes in mice, is and celiac disease, such posttranslationally modified proteins increased in islets from diabetes-prone nonobese diabetic behave as autoantigens (10,11), but their relevance in type 1 (NOD) mice. Finally, (pre)diabetic NOD mice have autoanti- – bodies and effector T cells that react against citrullinated diabetes is only starting to be explored (12 15). GRP78, indicating that inflammation-induced citrullination Building on our previous observation that the ER of GRP78 in b-cells generates a novel autoantigen in type 1 chaperone 78 kDa glucose-regulated protein (GRP78; also diabetes, opening new avenues for development named binding immunoglobulin protein [BiP]) is post- and therapeutic intervention. translationally modified in cytokine-exposed insulin- producing INS-1E cells (9), we now identify this modification as citrullination and show that inflammation-induced Type 1 diabetes is an autoimmune endocrine disease in citrullinated GRP78 is an autoantigen in type 1 diabetes. which loss of central and peripheral tolerance toward b-cell These findings suggest a novel role for GRP78 beyond its is proposed as the underlying mechanism. How- well-known function in the ER, leading to the loss of ever, b-cells themselves also contribute to trigger and/or tolerance to b-cells in type 1 diabetes. propagate the autoimmune attack, leading to a dialogue with immune infiltrating cells that may amplify local in- RESEARCH DESIGN AND METHODS flammation (insulitis) in genetically predisposed individuals Western blotting, , GRP78 cloning, (1). Insulin (or proinsulin) is probably the primary autoan- expression, purification, and in vitro citrullination are tigen in type 1 diabetes (2), but antigen spreading occurs as available in the Supplementary Data.

1Laboratory for Clinical and Experimental Endocrinology, KU Leuven, Leuven, This article contains Supplementary Data online at http://diabetes Belgium .diabetesjournals.org/lookup/suppl/doi:10.2337/db14-0621/-/DC1. 2 Department of Medical Protein Research, VIB, Ghent, Belgium D.R. and I.C. contributed equally to this study. 3Department of Biochemistry, Ghent University, Ghent, Belgium L.O. and C.M. contributed equally to this study. 4Laboratory for Cell Death Research and Therapy, KU Leuven, Leuven, Belgium 5Laboratory of Experimental Medicine and Université Libre de Bruxelles Center for © 2015 by the American Diabetes Association. Readers may use this article as Diabetes Research, Medical Faculty, Université Libre de Bruxelles, Brussels, Belgium long as the work is properly cited, the use is educational and not for profit, and the work is not altered. Corresponding author: Lut Overbergh, [email protected]. Received 18 April 2014 and accepted 28 August 2014. 574 GRP78 Is an Autoantigen in Type 1 Diabetes Diabetes Volume 64, February 2015

Reagents and Islet Isolation and Culture Primary antibodies were as follows: mouse anti-poly(ADP- Pancreatic islets were isolated from 3- or 10-week-old ) monoclonal (mAb) (Enzo Life Sciences, C57Bl/6, NOD, and NOR mice. Islet isolation and culture Antwerp, Belgium), rabbit anti-GRP78 and anti-CHOP were performed as previously described (16). C57Bl/6 polyclonal antibody (pAb) (Santa Cruz Biotechnology, islets were exposed to recombinant mouse IFNg (1,000 Santa Cruz, CA), rabbit anti-eIF2a pAb, anti–p-eIF2a units/mL) and recombinant human IL-1b (50 units/mL). (Ser51) pAb, anti-PERK mAb and anti–p-PERK(Thr980) Immunofluorescence mAb (Cell Signaling, Beverly, MA), and mouse anti-actin Immunofluorescence on isolated islets was performed as mAb (Sigma-Aldrich, Diegem, Belgium) for Western blot- ting; rabbit anti–b-catenin mAb (Cell Signaling Technol- previously described (17). Fixed INS-1E cells or sectioned islets were incubated with primary antibody in 1% BSA ogy) and mouse anti-GRP78 pAb (Abcam, Cambridge, for 1 h, followed by four washes in PBS before incubation U.K.) for immunocytochemistry. Anti–cit(510)-GRP78 with the secondary antibody in 1% BSA for another hour. was raised in rabbits against the following peptides: C- Nuclei were detected with DNA-binding dye DRAQ5TM aminohexanoic acid-IDVNGIL[]VTAEDKG- (Biostatus Ltd., Leicestershire, U.K.). Specificity was con- and acetyl-IDVNGIL[citrulline]VTAEDKG-aminohexanoic firmed by including negative controls with secondary anti- acid-C-amide, through five subsequent injections (21st fi bodies alone. INS-1E samples were observed under a Zeiss Century Biochemicals). Speci city of the antibody was fl 3 confirmed by dot blot against the citrullinated peptide LSM 510 microscope using a Plan-Neo uar 40 /1.3 oil DIC lens. Images were acquired and processed using LSM and its native counterpart. Secondary antibodies were as 510 software (Carl Zeiss AG, Jena, Germany). Mouse islet follows: donkey anti-rabbit horseradish peroxidase, don- sections were observed under a Nikon Eclipse Ti micro- key anti-mouse Alexa Fluor 488, and donkey anti-rabbit scope using a Plan-Fluor 403/0.75 DIC lens, and images Alexa Fluor 555 (Invitrogen, Merelbeke, Belgium). Oval- were acquired and processed using Nis-Elements Viewer bumin and rabbit peptidylarginine deiminase (PAD) en- 4.20 software (Nikon Instruments Inc.). zyme were from Sigma-Aldrich. Cell Surface Cell Lines and Culture Conditions INS-1E or MIN6 cells were incubated with the indicated Rat INS-1E cells, a gift from Prof. Wollheim (CMU, stressors for 12–15 h and then treated as previously de- Geneva, Switzerland), were cultured as previously de- scribed (18). scribed (9). INS-1E cells were exposed to recombinant rat IFNg (500 units/mL; R&D Systems), recombinant GRP78 ELISA human IL-1b (10 units/mL; R&D Systems), thapsigargin To determine GRP78 concentration in conditioned media (Tg) (15 and 50 nmol/L; Sigma-Aldrich), tunicamycin from INS-1E cells, the GRP78 ELISA kit (Enzo Life (Tn) (2 and 5 mg/mL; Sigma-Aldrich), high glucose Sciences, Antwerp, Belgium) was used, according to the (HG) (25 mmol/L; Sigma-Aldrich), and palmitate (Pa) manufacturer’s protocol. (0.5 mmol/L; Sigma-Aldrich). Mouse MIN6 cells, a gift Two-Dimensional Gel Electrophoresis Analysis from Dr. Miyazaki (Osaka University, Osaka, Japan), Two-dimensional gel electrophoresis (2D-GE) analysis was were cultured in DMEM (Invitrogen) containing 15% performed as previously described (9). (volume for volume) FCS, 100 units/mL penicillin, 100 mg/mL streptomycin, and 70 mmol/L b-mercaptoethanol. Quantitative RT-PCR MIN6 cells were exposed to recombinant mouse IFNg (500 Quantitative RT-PCR was performed as previously de- units/mL; R&D Systems) and human recombinant IL-1b scribed (19). (10 units/mL). Measurement of PAD Activity Measurements To determine PAD activity levels in islets and pancreata The percentage of living and apoptotic cells was assessed from C57Bl/6, NOR, and NOD mice, the antibody-based as previously described (9). assay for PAD activity (ABAP) (ModiQuest Research, Oss, the Netherlands) was used, according to the manufac- Mice turer’s protocol. C57Bl/6 mice were obtained from Harlan Laboratories (Horst, the Netherlands) and nonobese resistant (NOR) ELISA Against Native or Citrullinated mice from The Jackson Laboratory (Bar Harbor, ME). GRP78 Nonobese diabetic (NOD) mice have been inbred in our Serum autoantibodies against two native and citrullinated facility since 1989 and are kept under semi- GRP78 peptides (amino acids 500–519 TFEIDVNGILRV barrier conditions. For all experiments, a mix of male and TAEDKGTG and amino acids 295–314 AKRALSSQHQARI female mice was used. All animal manipulations were EIESFYE) were determined by ELISA as previously de- in compliance with the principles of laboratory care and scribed (20). For the citrullinated forms, arg510 or arg306 approved by the Institutional Animal Ethics Committee of were replaced by citrulline (synthesized by PolyPeptide Lab- KU Leuven. oratories, Strasbourg, France). diabetes.diabetesjournals.org Rondas and Associates 575

IFNg Measurement of ER stress to the observed cytokine-induced GRP78 mem- Splenocytes from 10-week-old and new-onset diabetic brane translocation by investigating the effect of the chem- NOD mice and age-matched C57Bl/6 and NOR mice were ical ER stressors Tg and Tn. Cytokines (12–15 h) induced cultured in flat-bottom 96-well plates (1 3 106 cells/well) a clear activation of the PERK-eIF2a-CHOP pathway, at the in the absence or presence of the indicated stimuli. IFNg dose tested (Supplementary Fig. 1). For Tg, there was levels were measured in cell culture supernatant after a dose-response effect both in terms of apoptosis (Fig. 48 h of culture, using Meso Scale Discovery technology 3A) and ER stress induction (Fig. 3B and Supplementary (Rockville, MD), according to the manufacturer’s protocol. Fig. 1), with minor ER stress at 15 nmol/L but a clear in- duction of the PERK-eIF2a-CHOP branch at 50 nmol/L Tg, Statistics toward levels similar to those observed upon cytokine ex- Statistical analyses of data were performed using GraphPad posure. Of note, Xbp1 splicing was induced by Tg at 50 Prism 6 (GraphPad Software, San Diego, CA). Data are nmol/L, which was not the case upon cytokine exposure expressed as means 6 SEM and were analyzed by a Kruskal- (Supplementary Fig. 1B). INS-1E cells were more resistant Wallis test followed by a Dunn multiple comparisons test, to Tn, with less apoptosis (Fig. 3A) and an intermediate unless stated otherwise in the figure legend. P values ,0.05 activation of PERK-eIF2a-CHOP and Xbp1 splicing, both at were considered significant. 2and5mg/mL (Supplementary Fig. 1), as compared with the higher doses of Tg and cytokines. As illustrated in Fig. RESULTS 3C–E,15nmol/LTgand2mg/mL Tn did not induce mem- GRP78 Is Translocated to the b-Cell Surface Upon brane translocation of GRP78. Importantly, the higher Cytokine Exposure dose of 50 nmol/L Tg led to a clear membrane transloca- We investigated cytokine-mediated regulation of GRP78 tion, which was paralleled by more marked expression of at both transcriptional and translational levels after 12–15 h ER stress markers (see above). A similar effect was ob- exposure and at cytokine concentrations that induced served with 5 mg/mL Tn, although less pronounced. only minor apoptosis (Fig. 1A). No significant increase Finally, upon metabolic stress (Pa [0.5 mmol/L] or the in GRP78 mRNA (Fig. 1B) and total protein expression combination of HG [25 mmol/L] + Pa), most of the ER (Fig. 1C, center panel, and Fig. 1D) were observed as stress markers were increased, although the increase in compared with control INS-1E cells. However, a detailed CHOP mRNA and protein was less marked than that analysis of the plasma membrane fraction revealed a lim- observed with chemical ER stressors or cytokines (Sup- ited presence of GRP78 at the cell surface under basal plementary Fig. 1). On the other hand, the total and conditions, which was increased upon IL-1b+IFNg expo- membrane-associated GRP78 protein levels remained un- F–H fi sure, but not upon single cytokine exposure (Fig. 1C, bot- altered (Fig. 3 ). Taken together, these ndings indi- tom panel, and Fig. 1E). This was confirmed in MIN6 cells cate that GRP78 membrane translocation occurs upon fl – after 12 h treatment (Fig. 1F), a preapoptotic time point in ammation- and severe chemical induced ER stress, (Fig. 1G). In parallel with the increased plasma membrane but not upon metabolic stress. translocation of GRP78, we also observed an increase in GRP78 Is Posttranslationally Modified in Cytokine- GRP78 secretion upon cytokine exposure of INS-1E cells Exposed b-Cells H (Fig. 1 ). Besides the above described membrane translocation of fi These observations were con rmed by immunocyto- GRP78, we observed extensive PTM of GRP78 upon chemistry showing increased GRP78 staining on the cytokine exposure of INS-1E cells (Fig. 4A and previously plasma membrane of nonpermeabilized cytokine-exposed described [9]) and C57Bl/6 mouse islets (Fig. 4B). This A INS-1E cells (Fig. 2 ), and a clear increase in colocaliza- was also the case, although to a lesser degree, for IL-1b tion between GRP78 and the plasma membrane marker or IFNg exposure alone (Fig. 4A). Of particular interest, B b-catenin in cytokine-exposed INS-1E cells (Fig. 2 ). Im- 2D-GE analysis of the plasma membrane fraction of con- portantly, a similar membrane translocation was observed trol and cytokine-exposed INS-1E (Fig. 4C) and MIN6 C in cytokine-exposed C57Bl/6 mouse islets (Fig. 2 ), again cells (Fig. 4D) not only demonstrated the presence of I at an early and preapoptotic time point (Fig. 1 ). Thus, the three different cytokine-responsive GRP78 isoforms, surface expression of GRP78 upon cytokine exposure is but also showed a cytokine-mediated upregulation of nu- not solely taking place in clonal b-cell models but also in merous acidic GRP78 isoforms as compared with control fi primary b-cells, and is not a consequence of nonspeci c cells. On the other hand, metabolic- and chemical-induced changes associated with cell death, increasing the rele- (both at low and high concentrations) ER stress did not fi vance of these ndings. induce detectable PTMs of GRP78 (Fig. 4E), confirming the specific effects of proinflammatory cytokines. GRP78 Is Translocated to the Plasma Membrane Upon Chemical ER Stress, but Not Upon Metabolic Stress GRP78 Is Citrullinated in Cytokine-Exposed INS-1E Cytokine exposure of INS-1E cells and primary rat, mouse, Cells and human b-cells is known to induce ER stress–dependent In order to identify the nature and site of cytokine- apoptosis (21–25). We further evaluated the contribution induced PTMs in GRP78, we subjected the three different 576 GRP78 Is an Autoantigen in Type 1 Diabetes Diabetes Volume 64, February 2015

Figure 1—Cytokine exposure induces membrane translocation of GRP78 in insulin-secreting INS-1E and MIN6 cells. A: Apoptosis levels in INS-1E cells exposed for 12–15 h (white bars) or 24 h (black bars) to IL-1b (10 units/mL) and/or IFNg (500 units/mL) (n =3–8 independent experiments, each biological replicate is the mean of two technical duplicates). B: GRP78 mRNA expression in INS-1E cells treated for 12–15 h as described above (n = 7, each biological replicate is the mean of two technical duplicates). C: Total and plasma membrane-associated (PM) GRP78 protein expression in INS-1E cells exposed to the indicated stressors. A representative Western blot from four independent experiments is shown. D and E: The relative intensities of the different protein bands were quantified by densitometry and expressed as a ratio (n =4).F: Total and PM GRP78 protein levels in control and cytokine-exposed MIN6 cells. A representative Western blot from two independent experiments is shown. G: Apoptosis levels in MIN6 cells exposed for 12 h (white bars) and 24 h (black bars) to IL-1b and IFNg (n =5–8 independent experiments, each biological replicate is the mean of two technical duplicates). H: GRP78 protein concentration in the culture medium of control and cytokine-exposed INS-1E cells. Data are expressed as means 6 SEM and were analyzed by a two-tailed paired Student t test (n = 10 independent experiments). I: Apoptosis levels in C57Bl/6 mouse islets exposed for 24 h (white bars) and 72 h (black bars) to IL-1b (50 units/mL) and IFNg (1,000 units/mL) (n =3–5 independent experiments, each biological replicate is the mean of two technical duplicates). *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 vs. respective control (Ctrl).

GRP78isoformsobservedinINS-1Ecellstomass digestion combined with differential N- spectrometry (MS) analysis. Sequence coverage ranged (Fig. 5A and B). As the peptide upstream of VTAEDKGTGNK from 62.54 to 74.46% (n = 2). When comparing the could not be identified using both methods, we hy- resulting peptide profiles of isoform 1 (I1) versus isoform pothesized the presence of a PTM in this region, possi- 2 (I2) and I1 versus isoform 3 (I3), one specific peptide, bly on Arg510, which would prevent tryptic digestion in VTAEDKGTGNK (AA511–521), was identified exclusively I2 and I3, thereby rendering the resulting peptide in I1. This was confirmed by a quantitative differential (AA493–521) too long for retrieval and detection by analysis using digestion in combination with differ- MS. Of note, also a second peptide of interest was ential N-butyrylation and endoproteinase Lys-C (endo Lys-C) found, which was identified almost exclusively in I2 and diabetes.diabetesjournals.org Rondas and Associates 577

Figure 2—Microscopic imaging of cytokine-induced GRP78 membrane translocation in INS-1E cells and mouse islets of Langerhans. A: Unpermeabilized control and cytokine-exposed INS-1E cells (15 h) were stained for GRP78 (green), and the nuclei (blue) were stained with Hoechst 33342. Images shown are representative for two independent experiments. Bar, 10 mm. Control and cytokine-exposed INS-1E cells (15 h) (B) and intact mouse islets (24 h) (C) were stained for GRP78 (green) and b-catenin (red). Nuclei (blue) were stained with Hoechst 33342. Images shown are representative for three independent experiments. Scale bars, 20 mm. 578 GRP78 Is an Autoantigen in Type 1 Diabetes Diabetes Volume 64, February 2015

Figure 3—Chemical ER stress, but not metabolic stress, induces membrane translocation of GRP78 in insulin-secreting INS-1E cells. A: Apoptosis levels in INS-1E cells exposed for 12–15 h (white bars) or 24 h (black bars) to Tg (15 or 50 nmol/L) or Tn (2 or 5 mg/mL) or HG (25 mmol/L), Pa (0.5 mmol/L), or the combination (HG+Pa) (n =5–9 independent experiments, each biological replicate is the mean of two technical duplicates). B: GRP78 mRNA expression in INS-1E cells treated for 12–15 h as described above (n =4–10, each biological replicate is the mean of two technical duplicates). C and F: Total and plasma membrane-associated (PM) GRP78 protein expression in INS- 1E cells exposed to the indicated stressors. A representative Western blot from four independent experiments is shown. D, E, G, and H: The relative intensities of the different protein bands were quantified by densitometry and expressed as a ratio (n =4–7). *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 vs. respective control (Ctrl).

I3, whereas hardly in I1 (AA307–324) (Fig. 5A and B). isoelectric point without change in molecular weight and Since this would point to the presence of a PTM in the described to occur in GRP78 in other cell types, namely basic but not acidic forms, we did not further analyze the ADP ribosylation, , and citrullination relevance of this peptide. (26–29). Both ADP ribosylation and citrullination occur Based on these findings, we then investigated the on Arg residues, whereas phosphorylation may occur on nature of PTM present in the identified region. We did Thr 518. not succeed in retrieving by MS the longer, Arg510- We initially investigated ADP ribosylation and phos- containing peptide (AA493–521) or a spiked tryptic pep- phorylation of GRP78, but the absence of a positive signal tide with heavy label (AA493–521) used as internal control. in 2D Western blots of cytokine-exposed INS-1E cells with This is probably caused by insolubility and inability to anti–poly(ADP-ribose) (Fig. 6A) and the absence of GRP78 analyze on a C18 column and forced us to use instead staining by Pro-Q Diamond in control- and cytokine- specific enzymatic and antibody-based assays. We focused exposed INS-1E cells (Fig. 6B), as well as persistence of the on three potential PTMs consistent with an acidic shift in modified isoforms upon treatment with calf intestinal diabetes.diabetesjournals.org Rondas and Associates 579

Figure 4—Cytokine exposure induces PTM of GRP78 in INS-1E cells and in intact mouse islets. Representative images from 2D difference gel electrophoresis analysis (selected region of 24 cm, pH 4–7, 12.5% SDS-PAGE) of GRP78 in control and cytokine-exposed INS-1E cells (A) and intact mouse islets (B) with corresponding three-dimensional view and Graph view of the DeCyder analysis. For each of the three isoforms (I1, I2, and I3), the fold increase or decrease is shown, and statistical analysis was performed using a two-tailed, unpaired Student t test (n = 4 independent experiments; *P < 0.05 and **P < 0.01 vs. control). Representative images of 2D-GE analysis (selected region of 24 cm, pH 4–7, 12.5% SDS-PAGE) of intracellular and membrane-associated GRP78 in control and cytokine-exposed INS-1E cells (one representative experiment out of five independent experiments is shown) (C) and MIN6 cells (one representative experiment out of two independent experiments is shown) (D). E: Representative 2D-GE analysis (selected region of 24 cm, pH 4–7, 12.5% SDS-PAGE) with corresponding three-dimensional view of GRP78 in total lysates from control and differentially exposed (as indicated) INS-1E cells (one representative experiment out of three independent experiments is shown). 580 GRP78 Is an Autoantigen in Type 1 Diabetes Diabetes Volume 64, February 2015

Figure 5—Quantitative mass spectrometric analysis of the three GRP78 isoforms I1, I2, and I3 using trypsin digestion combined with differential N-butyrylation and endo Lys-C digestion combined with differential N-propionylation. For each dataset, the ratios (light [I1]/ heavy [I2 or I3]) were converted to their log2 value, in order to render a normal distribution. In the volcano plots, the size of the fold change is compared with the statistical significance level. Comparison between the most basic isoform of GRP78 (I1) and the first more acidic isoform (I2) (A) and between I1 and the second more acidic GRP78 isoform (I3) (B) with a sequence coverage of 77.67 and 71.70%, respectively (494 and 456 out of 654 amino acids of the GRP78_RAT sequence, respectively, with omission of the 18– signal peptide). The identified tryptic peptides are marked in red, the identified endo Lys-C peptides in bold, and the signal peptides in blue. The two peptides with the highest z score and fold change are depicted on the volcano plot and underlined in the sequence.

alkaline phosphatase (Fig. 6C)orl-phosphatase (data not which were either low or undetectable. Further analysis shown), argued against both modifications. in other tissues revealed an overall very low expression To verify the implication of citrullination, an antibody of Padi2 in the immune-related tissues thymus, lymph that specifically recognizes citrullinated GRP78 at Arg510 nodes, and spleen. Except for kidney, no elevated levels (selection based on the obtained MS data, see Fig. 5) was of Padi2 were observed in the other tissues analyzed in raised in rabbits. 2D Western blots from control and NOD as compared with C57Bl/6 mice. In addition, NOR cytokine-exposed INS-1E cells with anti–cit510-GRP78 mice showed even lower/undetectable Padi2 expression clearly indicated that the cytokine-induced acidic isoforms in most of the tissues analyzed (Fig. 7C). Furthermore, of GRP78 correspond to citrullinated GRP78 at residue elevated Padi2 mRNA expression in NOD islets corre- Arg510, whereas no reactivity was observed against the sponded to higher PAD activity in total pancreases most basic, nonmodified GRP78 isoform 1 (Fig. 6D). (Fig. 7D and E)andislets(Fig.7F and G) of 3- and 10- week-old NOD mice, compared with both C57Bl/6 and Padi2 Expression and Activity Are Upregulated in NOR mice, indicating a marked increase of PAD activity NOD Mice in islets of NOD mice immediately before and during Next, we investigated the potential role of citrullination insulitis. Of note, 3-week-old NOD mice did not show in the diabetes-prone NOD mouse. Elevated Padi2 mRNA any sign of inflammation in the islets, as measured by IL- expression was observed in islets of 3- and 10-week-old 1b (Fig. 7H)andIFNg (Fig. 7I) mRNA expression. In 10- prediabetic NOD mice as compared with islets from week-old NOD mice, on the other hand, clear signs of age-matchedC57Bl/6andNORcontrolmice(Fig.7A immune infiltration were observed, as evidenced by high and B). No differences between the strains were ob- expression of IFNg and IL-1b mRNA, confirming previ- served for Padi1, Padi3, Padi4,andPadi6 expression, ous findings from our group (30). diabetes.diabetesjournals.org Rondas and Associates 581

Figure 6—GRP78 is citrullinated upon cytokine exposure of INS-1E cells. A: Representative 2D Western blot (11 cm, pH 4–7, 4–12.5% SDS-PAGE) of cytokine-exposed INS-1E lysate detected with anti–poly(ADP-ribose) antibody (top panel) followed by anti-GRP78 (bottom panel) (one representative experiment out of three independent experiments is shown). B: 2D-GE gel of cytokine-exposed INS-1E cells (24 cm, pH 4–7, 12.5% SDS-PAGE) stained using Sypro Ruby to visualize all proteins (top panel) and Pro-Q Diamond dye to detect phos- phoproteins (bottom panel) (one representative experiment out of two independent experiments is shown). C: 2D-GE analysis of GRP78 in control and cytokine-exposed INS-1E cells (selected region of pH 4–7, 12.5% SDS-PAGE is shown), treated or not with calf intestinal alkaline phosphatase (CIAP) (one representative experiment out of two independent experiments is shown). D: 2D-GE analysis of citrullinated GRP78 in control and cytokine-exposed INS-1E cells with anti–cit510-GRP78 (top panels) and total GRP78 Ab (bottom panels) (one representative experiment out of seven independent experiments is shown).

NOD Mice Have Circulating Autoantibodies and vitro citrullinated recombinant mouse GRP78 protein. Autoreactive T Cells Against Citrullinated GRP78 Secretion of IFNg wasusedasameasureofeffectorT-cell To evaluate whether citrullinated GRP78 contributes to the activation. Whereas little to no effector T-cell activation was autoimmune response in NOD mice, serum samples from observed in the three different strains when culturing prediabetic and new-onset diabetic NOD mice and age- splenocytes with different concentrations of native matched C57Bl/6 and NOR mice were analyzed for the pres- GRP78 (Fig. 8B, red line and top right graph), a clear, ence of autoantibodies against the native and citrullinated dose-responsive increase in IFNg secretion was observed peptide containing the epitope of interest (p500–519). when splenocytes from prediabetic and diabetic NOD mice Serum levels of anti-GRP78 antibodies recognizing this were cultured in the presence of citrullinated GRP78 (Fig. citrullinated epitope were significantly higher in new-onset 8B, blue line and bottom right graph). C57Bl/6 splenocytes diabetic NOD mice as compared with age-matched C57Bl/6 were unresponsive to citrullinated GRP78, whereas a minor or NOR mice (Fig. 8A, right panel). Furthermore, in diabetic IFNg response was detected in NOR splenocytes. Absence NOD mice, significant higher serum antibody levels to the of an IFNg response against both the PAD enzyme alone citrullinated peptide were found compared with those of and the control protein ovalbumin, either native or in vitro the native peptide. No such differences were observed in citrullinated (at 0.1, 1, and 5 mg/mL) (data not shown), the case of another irrelevant (citrullinated) GRP78 peptide suggests that the observed autoreactive T-cell response is (p295–314) tested (Supplementary Fig. 2). These findings specifically generated against citrullinated GRP78. provide evidence that this specific citrullinated epitope is important for autoantibody generation during type 1 diabetes DISCUSSION development in NOD mice. The role of posttranslationally modified proteins is well Next, to determine if NOD mice have autoreactive established in several human autoimmune diseases, and T cells against native or citrullinated GRP78, freshly isolated evidence for similar phenomena in the development of splenocytes from prediabetic and new-onset diabetic NOD type 1 diabetes is accumulating (12–14). Most impor- mice and age-matched C57Bl/6 and NOR mice were tantly for this study, McGinty et al. (15) recently demon- stimulated with various concentrations of native and in strated the relevance of citrullination in patients with 582 GRP78 Is an Autoantigen in Type 1 Diabetes Diabetes Volume 64, February 2015

Figure 7—NOD islets have high Padi2 mRNA expression and PAD activity. A and B: Padi1, 2, 3, 4, and 6 mRNA expression in islets of Langerhans of respectively 3- and 10-week-old C57Bl/6 (black bars), NOR (gray bars), and NOD (white bars) mice (n =5–8). C: Padi2 mRNA levels in different tissues of 3-week-old C57Bl/6 (black bars), NOR (gray bars), and NOD (white bars) mice (n =5–10, each sample consists of tissue isolated from a single mouse). D and E: Pancreatic PAD activity in respectively 3- and 10-week-old C57Bl/6 (black bars), NOR (gray bars), and NOD (white bars) mice. Data are expressed as means 6 SEM and were analyzed by a one-way ANOVA followed by a Bonferroni multiple comparisons test (n =4–9). F and G: Islet PAD activity in respectively 3- and 10-week-old C57Bl/6 (black bars), NOR (gray bars), and NOD (white bars) mice (n = 4). All replicates refer to biological replicates with samples (islets or pancreas) obtained from individual mice. The PAD activity experiments were performed at least three times. ND, not detectable. IL-1b (H) and IFNg (I) mRNA levels in 3- and 10-week-old C57Bl/6 (black bars), NOR (gray bars), and NOD (white bars) mice (n =5–8, each sample consists of islets isolated from a single mouse). *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

type 1 diabetes, by showing an increased response to direct upregulation of the PADI2 enzyme by cytokines citrullinated GAD65 peptides. We show that the ER chap- (data not shown) but needs to be the consequence of erone GRP78 is citrullinated specifically upon exposure of increased PAD activity in b-cells exposed to cytokines. b-cells to inflammatory stress. This is paralleled by trans- Since PAD activity is highly Ca2+ dependent, changes in Ca2+ location of GRP78 to the b-cell plasma membrane and fluxes induced in cytokine-exposed b-cells might play a role. eventually its secretion. Under these circumstances, a spe- The present findings, together with the recent report on cific cross-talk between the b-cell and the immune system posttranslationally modified GAD65 (15), add type 1 diabetes is initiated, resulting in the generation of autoantibodies to the list of autoimmune diseases involving citrullination, and induction of T-cell autoreactivity against citrullinated i.e., RA, multiple sclerosis, and systemic erhythematosus (32). GRP78 (Fig. 8C). Importantly, we also observed a marked This suggests that citrullination is not a specificdisease- upregulation of Padi2 in islets of NOD mice, providing related event but rather an inflammation-dependent pro- a strong argument for PADI2 being the diabetes suscep- cess occurring preferentially in autoimmune target tissues. tibility gene in the recently identified Idd25 locus on The role of citrullination in the induction of auto- mouse chromosome 4 (31) and adding to a potential antigenicity has been best described in RA, with several role for citrullination in type 1 diabetes (15). citrullinated autoantigens, including GRP78, already iden- In previous studies, we have shown that GRP78 is tified (20,33). These citrullinated peptide epitopes are bet- posttranslationally modified in INS-1E cells exposed to ter accommodated in the HLA pocket of HLA-DR4–type cytokines (9), a PTM that we now identified as citrullina- individuals, determining the strength of the immune re- tion. Cytokines contribute to b-cell dysfunction and death sponse to citrullinated peptides and providing a molecular at least in part through inducing ER stress (23–25). How- basis for the genetic predisposition of HLA-DR4 individuals ever, citrullination of GRP78 is not induced by chemical to RA (34). This mechanism has recently also been de- ER stressors (Tg or Tn) or by metabolic stress via expo- scribed in type 1 diabetes (15), where HLA-DR4 is an im- sure to HG and/or Pa, suggesting that cytokine-induced portant risk allele for the disease (35). GRP78 citrullination occurs through a mechanism inde- In addition to citrullination, cytokine-induced trans- pendent of its ER stress–inducing capacities. Preliminary location of GRP78 to the b-cell plasma membrane may be data suggest that this citrullination is not mediated by a crucial step for GRP78 to become an autoantigen. This diabetes.diabetesjournals.org Rondas and Associates 583

Figure 8—NOD mice have circulating autoantibodies and autoreactive T cells against citrullinated GRP78. A: Comparison between the serum levels of antinative and anticitrullinated peptide (AA500–519) antibodies in prediabetic and diabetic NOD (NOD DM) mice and age- matched C57Bl/6 and NOR mice grouped according to age. Each dot indicates the value of a single mouse. Four independent experiments were performed, each containing samples of all experimental groups. B: IFNg response of splenocytes from prediabetic (n = 9, light gray bars) and diabetic NOD (n = 11, white bars) mice and age-matched C57Bl/6 (n = 8, black bars) and NOR (n = 11, dark gray bars) mice stimulated with various concentrations of native (red) and citrullinated (blue) recombinant mouse GRP78. Four independent experiments were performed, with two to three per group per experiment. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. C: Proposed model for the role of b-cell citrullinated GRP78 in autoantibody generation and T-cell activation. Exposure of b-cells to cytokines induces citrullination, membrane translocation, and secretion of GRP78. Citrullinated membrane–associated or secreted GRP78 is taken up and processed by B cells and antigen-presenting cells (APCs), resulting in the generation of specific anticitrullinated GRP78 autoantibodies and release of IFNg by activated effector T cells, respectively. 584 GRP78 Is an Autoantigen in Type 1 Diabetes Diabetes Volume 64, February 2015 translocation is shown to be an early event in response to Whether these observed processes are also applicable inflammation, suggesting that it is an active process at to other ER chaperones or heat shock proteins, such as least in part independent from “protein leakage” by apo- HSP60, requires further investigation. It will be of utmost ptotic b-cells. GRP78 membrane translocation, but not importance to determine whether similar processes citrullination, was found to be ER stress dependent and are involved in human type 1 diabetes. The knowledge paralleled to the increased CHOP expression. Membrane- that inflammation-mediated b-cellstressistakingplace associated GRP78 has been described in different tumor cell in human type 1 diabetes (48) and the high overlap types (36–38) as well as in exocrine pancreatic cells (39) and between autoantigens identified to date in NOD mice proliferating endothelial and monocytic cells (40,41). In and type 1 diabetes patients (49) are in support of this these models, membrane-associated GRP78 acts, among hypothesis. other functions, as a cell surface signaling receptor for dif- In conclusion, local inflammation in the islets induces ferent ligands such as activated a2-macroglobulin (42,43) citrullination of GRP78 in the stressed b-cells, turning and coxsackie A9 virus (44,45) and is found associated citrullinated GRP78 into an autoantigen. This modified with the major histocompatibility complex class I (MHC-I) GRP78 is recognized by both B and T cells, thus propa- (44). Of interest, changes in the topography of membrane- gating and amplifying the ongoing autoimmune attack associated GRP78, caused by PTMs, may convert GRP78 against the b-cells. Our findings support and provide into a receptor with autoantigenic properties. This has mechanistic evidence for the concept that inflammation- been observed in cancer cells where GRP78 is modified by induced b-cell stress initiates a specificcommunication O-linked (38), leading to the generation of between the b-cell and the immune system, which will GRP78 autoantibodies. The present observations that in- aggravate and accelerate the development of type 1 dia- flammation induces extensive GRP78 citrullination, trans- betes. A genetic predisposition for increased citrullination location to the plasma membrane, and secretion underscore in the islets, as shown here for NOD mice, is expected to its putative function as an autoantigen in type 1 diabetes. further exacerbate this process. This proposed mechanism Furthermore, based on the proposed transmembrane (Fig. 8C), implicating tissue-specificandinflammation- model for GRP78 (46), the reactive p500–519 epitope, induced protein modification, cell surface translocation, against which GRP78 autoantibodies are generated in and secretion, would also explain why different tissue- NOD mice, is located in the extracellular , thus specific autoimmune diseases can have similar autoantigens. being exposed to infiltrating immune cells. The possible role for citrullinated GRP78 as an auto- antigen in type 1 diabetes is supported by the present Acknowledgments. The technical assistance of Frea Coun, Martine observations showing the generation of autoantibodies as Gilis, Jos Laureys, Willem Van Den Berghe, Wim Werckx, and Farah-Deborah Lok (Laboratory of Clinical and Experimental Endocrinology, KU Leuven) is greatly well as CD4+ T-cell autoreactivity against citrullinated appreciated. The authors thank Katleen Lemaire (Gene Expression Group, KU GRP78 in NOD mice. No T-cell reactivity is observed in Leuven) and Monique Beullens (Laboratory for Biosignaling and Therapeutics, KU C57Bl/6 mice, whereas NOR mice show a minor T-cell re- Leuven) for advice on GRP78 cloning in pET vector. The authors thank the Cell fl sponse. This supports the idea that in ammation is neces- Imaging Core (KU Leuven) for providing technical assistance with confocal sary to initiate this process, as low levels of IFNg and IL-1b microscopy. are detected in islets of 10-week-old NOR mice, a phenom- Funding. This work was supported by Juvenile Diabetes Research Foundation enon referred to as protracted, noninvasive insulitis (47). A International (17-2012-129 and 17-2013-515), the European Community’s Sev- marked upregulation of PAD2, a key enzyme for protein enth Framework Programme NAIMIT (Natural Immunomodulators as Novel citrullination, is observed exclusively in islets from NOD Immunotherapies for Type 1 Diabetes) under grant agreement 241447, the mice, as compared with NOR and C57Bl/6 islets. PAD ac- KU Leuven (Geconcerteerde Onderzoeksactie GOA 12/24 and an F+ fellowship tivity is further increased in NOD islets with increasing age for D.R.), and the Flemish Research Foundation (G.0619.12, a postdoctoral fel- lowship for G.B.F. and W.D. and a clinical research fellowship for C.M.). and insulitis, suggesting that inflammation plays a role for 2+ Duality of Interest. No potential conflicts of interest relevant to this article this phenomenon, perhaps by increasing cytosolic Ca are to be reported. concentrations due to cytokine-mediated calcium depletion Author Contributions. D.R. and I.C. designed and performed research, Padi2 from the ER (24). Interestingly, expression of was analyzed data, and wrote the manuscript. W.D., G.B.F., A.S., and A.D.G. per- very low in NOD thymus and not different from C57Bl/6 and formed research. D.L.E. analyzed data and wrote and edited the manuscript. P.A. NOR thymus. This may explain the escape of citrullinated and K.G. edited the manuscript. L.O. and C.M. designed research and wrote and GRP78 from thymic tolerization in developing thymo- edited the manuscript. L.O. and C.M. are the guarantors of this work and, as cytes, thereby clarifying why citrullinated GRP78 can be such, had full access to all the data in the study and take responsibility for the recognized as a b-cell–specific autoantigen whereas native integrity of the data and the accuracy of the data analysis. GRP78 is ubiquitously expressed. A similar mechanism References has been proposed for chromogranin A, where exposure 1. Eizirik DL, Colli ML, Ortis F. The role of inflammation in insulitis and beta- of a naturally occurring cleavage product (peptide WE14) cell loss in type 1 diabetes. Nat Rev Endocrinol 2009;5:219–226 to , expressed in b-cells, forms high- and 2. Daniel D, Gill RG, Schloot N, Wegmann D. Epitope specificity, cytokine low-molecular weight aggregates, thus rendering the pep- production profile and diabetogenic activity of insulin-specific T cell clones iso- tide highly antigenic in NOD mice (13). lated from NOD mice. Eur J Immunol 1995;25:1056–1062 diabetes.diabetesjournals.org Rondas and Associates 585

3. Baekkeskov S, Aanstoot HJ, Christgau S, et al. Identification of the 64K reticulum Ca2+, leading to induction of endoplasmic reticulum stress in pan- autoantigen in insulin-dependent diabetes as the GABA-synthesizing enzyme creatic beta-cells. Diabetes 2005;54:452–461 decarboxylase [published correction appears in Nature 1990;347: 25. Allagnat F, Fukaya M, Nogueira TC, et al. C/EBP homologous protein con- 782]. Nature 1990;347:151–156 tributes to cytokine-induced pro-inflammatory responses and apoptosis in 4. Kawasaki E, Eisenbarth GS, Wasmeier C, Hutton JC. Autoantibodies to b-cells. Cell Death Differ 2012;19:1836–1846 protein phosphatase-like proteins in type I diabetes. Overlapping spe- 26. Chambers JE, Petrova K, Tomba G, Vendruscolo M, Ron D. ADP ribosylation cificities to phogrin and ICA512/IA-2. Diabetes 1996;45:1344–1349 adapts an ER chaperone response to short-term fluctuations in unfolded protein 5. Brudzynski K, Martinez V, Gupta RS. Secretory granule autoantigen in load. J Cell Biol 2012;198:371–385 insulin-dependent diabetes mellitus is related to 62 kDa heat-shock protein 27. Nakai A, Kawatani T, Ohi S, et al. Expression and phosphorylation of BiP/ (hsp60). J Autoimmun 1992;5:453–463 GRP78, a molecular chaperone in the endoplasmic reticulum, during the 6. Stadinski BD, Delong T, Reisdorph N, et al. Chromogranin A is an auto- differentiation of a mouse myeloblastic cell line. Cell Struct Funct 1995;20: antigen in type 1 diabetes. Nat Immunol 2010;11:225–231 33–39 7. Ortis F, Naamane N, Flamez D, et al. Cytokines interleukin-1beta and tumor 28. Lu MC, Lai NS, Yu HC, Huang HB, Hsieh SC, Yu CL. Anti-citrullinated protein necrosis factor-alpha regulate different transcriptional and alternative splicing antibodies bind surface-expressed citrullinated Grp78 on monocyte/macrophages networks in primary beta-cells. Diabetes 2010;59:358–374 and stimulate tumor necrosis factor alpha production. Arthritis Rheum 2010;62: 8. D’Hertog W, Maris M, Ferreira GB, et al. Novel insights into the global 1213–1223 proteome responses of insulin-producing INS-1E cells to different degrees of 29. Laitusis AL, Brostrom MA, Brostrom CO. The dynamic role of GRP78/BiP in endoplasmic reticulum stress. J Proteome Res 2010;9:5142–5152 the coordination of mRNA with protein processing. J Biol Chem 1999; 9. D’Hertog W, Overbergh L, Lage K, et al. analysis of cytokine- 274:486–493 induced dysfunction and death in insulin-producing INS-1E cells: new insights 30. Jörns A, Arndt T, Meyer zu Vilsendorf A, et al. Islet infiltration, cytokine into the pathways involved. Mol Cell Proteomics 2007;6:2180–2199 expression and beta cell death in the NOD mouse, BB rat, Komeda rat, 10. Bicker KL, Thompson PR. The protein deiminases: Structure, LEW.1AR1-iddm rat and humans with type 1 diabetes. Diabetologia 2014;57: function, inhibition, and disease. Biopolymers 2013;99:155–163 512–521 11. Sollid LM, Jabri B. Celiac disease and transglutaminase 2: a model for 31. Stolp J, Chen YG, Cox SL, et al. Subcongenic analyses reveal complex posttranslational modification of antigens and HLA association in the pathogen- interactions between distal chromosome 4 genes controlling diabetogenic B cells esis of autoimmune disorders. Curr Opin Immunol 2011;23:732–738 and CD4 T cells in nonobese diabetic mice. J Immunol 2012;189:1406–1417 12. Dunne JL, Overbergh L, Purcell AW, Mathieu C. Posttranslational mod- 32. Chirivi RGS, van Rosmalen JWG, Jenniskens GJ, Pruijn GJ, Raats JMH. ifications of proteins in type 1 diabetes: the next step in finding the cure? Di- Citrullination: a target for disease intervention in multiple sclerosis and other abetes 2012;61:1907–1914 inflammatory diseases? J Clin Cell Immunol 2013;4:146 13. Delong T, Baker RL, He J, Barbour G, Bradley B, Haskins K. Diabetogenic T-cell 33. Bläss S, Union A, Raymackers J, et al. The stress protein BiP is overex- clones recognize an altered peptide of chromogranin A. Diabetes 2012;61:3239–3246 pressed and is a major B and T cell target in rheumatoid arthritis. Arthritis Rheum 14. van Lummel M, Duinkerken G, van Veelen PA, et al. Posttranslational 2001;44:761–771 modification Of HLA-DQ binding islet autoantigens in type 1 diabetes. Diabetes 34. Scally SW, Petersen J, Law SC, et al. A molecular basis for the association 2014;63:237–247 of the HLA-DRB1 locus, citrullination, and rheumatoid arthritis. J Exp Med 2013; 15. McGinty JW, Chow IT, Greenbaum C, Odegard J, Kwok WW, James EA. 210:2569–2582 Recognition of posttranslationally modified GAD65 epitopes in subjects with type 35. Noble JA, Valdes AM. Genetics of the HLA region in the prediction of type 1 1 diabetes. Diabetes 2014;63:3033–3040 diabetes. Curr Diab Rep 2011;11:533–542 16. Gysemans CA, Waer M, Valckx D, et al. Early graft failure of xenogeneic islets 36. Mintz PJ, Kim J, Do KA, et al. Fingerprinting the circulating repertoire of in NOD mice is accompanied by high levels of interleukin-1 and low levels of antibodies from cancer patients. Nat Biotechnol 2003;21:57–63 transforming growth factor-beta mRNA in the grafts. Diabetes 2000;49:1992–1997 37. Chinni SR, Falchetto R, Gercel-Taylor C, Shabanowitz J, Hunt DF, Taylor DD. 17. Rondas D, Tomas A, Halban PA. Focal adhesion remodeling is crucial for Humoral immune responses to cathepsin D and glucose-regulated protein 78 in glucose-stimulated insulin secretion and involves activation of focal adhesion ovarian cancer patients. Clin Cancer Res 1997;3:1557–1564 kinase and paxillin. Diabetes 2011;60:1146–1157 38. Rauschert N, Brändlein S, Holzinger E, Hensel F, Müller-Hermelink HK, 18. Garg AD, Dudek AM, Ferreira GB, et al. ROS-induced autophagy in cancer Vollmers HP. A new tumor-specific variant of GRP78 as target for antibody-based cells assists in evasion from determinants of immunogenic cell death. Autophagy therapy. Lab Invest 2008;88:375–386 2013;9:1292–1307 39. Takemoto H, Yoshimori T, Yamamoto A, et al. Heavy chain binding protein 19. Maris M, Robert S, Waelkens E, et al. Role of the saturated nonesterified (BiP/GRP78) and endoplasmin are exported from the endoplasmic reticulum in rat fatty acid palmitate in beta cell dysfunction. J Proteome Res 2013;12:347– exocrine pancreatic cells, similar to protein disulfide-isomerase. Arch Biochem 362 Biophys 1992;296:129–136 20. Shoda H, Fujio K, Shibuya M, et al. Detection of autoantibodies to citrullinated 40. Davidson DJ, Haskell C, Majest S, et al. Kringle 5 of human plasminogen BiP in rheumatoid arthritis patients and pro-inflammatory role of citrullinated BiP induces apoptosis of endothelial and tumor cells through surface-expressed in collagen-induced arthritis. Arthritis Res Ther 2011;13:R191 glucose-regulated protein 78. Cancer Res 2005;65:4663–4672 21. Rondas D, Bugliani M, D’Hertog W, et al. Glucagon-like peptide-1 protects 41. Bhattacharjee G, Ahamed J, Pedersen B, et al. Regulation of tissue factor— human islets against cytokine-mediated b-cell dysfunction and death: a proteo- mediated initiation of the coagulation cascade by cell surface grp78. Arterioscler mic study of the pathways involved. J Proteome Res 2013;12:4193–4206 Thromb Vasc Biol 2005;25:1737–1743 22. Eizirik DL, Cardozo AK, Cnop M. The role for endoplasmic reticulum stress 42. Misra UK, Chu CT, Gawdi G, Pizzo SV. Evidence for a second alpha 2- in diabetes mellitus. Endocr Rev 2008;29:42–61 macroglobulin receptor. J Biol Chem 1994;269:12541–12547 23. Engin F, Yermalovich A, Nguyen T, et al. Restoration of the unfolded protein 43. Misra UK, Gonzalez-Gronow M, Gawdi G, Hart JP, Johnson CE, Pizzo SV. response in pancreatic beta cells protects mice against type 1 diabetes. Sci The role of Grp 78 in alpha 2-macroglobulin-induced signal transduction. Evi- Transl Med 2013;5:211ra156 dence from RNA interference that the low density lipoprotein receptor-related 24. Cardozo AK, Ortis F, Storling J, et al. Cytokines downregulate the sarco- protein is associated with, but not necessary for, GRP 78-mediated signal endoplasmic reticulum pump Ca2+ ATPase 2b and deplete endoplasmic transduction. J Biol Chem 2002;277:42082–42087 586 GRP78 Is an Autoantigen in Type 1 Diabetes Diabetes Volume 64, February 2015

44. Triantafilou M, Fradelizi D, Triantafilou K. Major histocompatibility class one 47. Fox CJ, Danska JS. Independent genetic regulation of T-cell and antigen- associates with glucose regulated protein (GRP) 78 on the cell surface. presenting cell participation in autoimmune islet inflammation. Diabetes 1998;47: Hum Immunol 2001;62:764–770 331–338 45. Triantafilou K, Fradelizi D, Wilson K, Triantafilou M. GRP78, a coreceptor for 48. Marhfour I, Lopez XM, Lefkaditis D, et al. Expression of endoplasmic re- coxsackievirus A9, interacts with major histocompatibility complex class I mol- ticulum stress markers in the islets of patients with type 1 diabetes. Diabetologia ecules which mediate virus internalization. J Virol 2002;76:633–643 2012;55:2417–2420 46. Gonzalez-Gronow M, Selim MA, Papalas J, Pizzo SV. GRP78: a multifunc- 49. Chaparro RJ, Dilorenzo TP. An update on the use of NOD mice to study tional receptor on the cell surface. Antioxid Redox Signal 2009;11:2299–2306 autoimmune (type 1) diabetes. Expert Rev Clin Immunol 2010;6:939–955